PROJECT SUMMARY Neuropsychiatric symptoms (NPS) are core features of Alzheimer's disease (AD) and related dementias that are associated with major adverse effects on daily function and quality of life, and accelerate time to institutionalization. Of all the NPS, depression is the most frequently observed symptom in people with mild cognitive impairment and early AD. As the disease progresses, agitation, delusions and hallucinations become more common, whereas apathy is the most persistent and frequent NPS throughout all the stages of AD. AD-NPS share some clinical features with serious mental illnesses (SMIs), such as schizophrenia, bipolar disorder and major depressive disorder, but whether these conditions share similar aethiopathies is unclear. Given that reliable treatments for NPS in the context of AD and other dementias do not exist, a better understanding of the molecular mechanisms and pathways underlying NPS in AD and other neuropsychiatric illnesses is a critical next step to identify reliable biomarkers that could lead to novel therapeutics. There are two overarching goals of this proposal. First, we will identify the molecular mechanisms and neuropathological changes that are associated with the presence of NPS in patients with AD. Second, we will examine if the mechanisms of pathology associated with NPS are shared or distinct among AD and SMIs. More specifically, we propose to build multi-scale integrative models using phenomics and genomics data from 1,264 autopsy cases derived from a single brain bank. The bank includes detailed phenomics data such as well characterized NPS, clinical diagnosis (AD and other neurodegenerative or neuropsychiatric traits), severity of cognitive decline and neuropathology for each patient sample. From each case, we will apply innovative approaches that reduce the cost and technical biases associated with conventional methods, and capture gene expression signatures and epigenetic regulatory elements at the single-cell level. Novel deep-learning methods will be applied for the multi-scale integration of neuropathologic changes with genetic markers and functional genomic changes (such as changes in gene expression and enhancer sequences) within specific cell types, to predict various NPS in AD and other neuropsychiatric traits; we refer to these integrative models as genotype- marker-phenotype models. We expect that these models will enable us to assign genotypes and molecular markers to specific NPS within AD and other neuropsychiatric traits at the single-cell level, an unprecedented level of resolution. In addition, we will test the translational potential of the genotype-marker-phenotype models to predict AD-NPS using independent large-scale biobank datasets, in which genotypes and electronic health records are available. Successful completion of the proposed studies will have immediate utility by generating potential biomarkers for NPS diagnosis and prognosis and by providing predictive models for patient stratification in clinical trials. In the longer term, our models will help us create a blueprint for therapeutic strategies and interventions to treat NPS in AD.